Add GMM-Demux

modules/nf-core/gmmdemux/environment.yml

@@ -0,0 +1,9 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/environment-schema.json
+name: "gmmdemux"
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+dependencies:
+  - "bioconda::gmm-demux=0.2.2.3"

modules/nf-core/gmmdemux/main.nf

@@ -0,0 +1,71 @@
+
+process GMMDEMUX {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/gmm-demux:0.2.2.3--pyh7cba7a3_0':
+        'biocontainers/gmm-demux:0.2.2.3--pyh7cba7a3_0' }"
+
+    input:
+    tuple val(meta), path(cell_hashing_barcodes,stageAs: "hto_files/*"), path(cell_hashing_matrix,stageAs: "hto_files/*"),path(cell_hashing_features,stageAs: "hto_files/*"),val(hto_names)
+    val csv
+    val num_cells
+    val output_dir
+    val full_report
+    val simplified_report
+    val examine
+    val ambigous
+    val extract
+    val skip
+
+    output:
+    tuple val(meta), path('test/barcodes.tsv.gz'), emit: barcodes
+    tuple val(meta), path('test/*.mtx.gz'), emit: matrix
+    tuple val(meta), path('test/features.tsv.gz'), emit: features
+    path "versions.yml"           , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    //since this tool has many optional inputs that can be passed in, we need to check if they are null or not
+    // in order to produce or not certain reports or outputs
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    def n_cells = num_cells ? "--summary $num_cells" : ""
+    def output_path = output_dir ? "-o $prefix" : ""
+    def full_rep = full_report ? "-f $prefix" : ""
+    def simplified_rep = simplified_report ? "--simplified $simplified_report" : ""
+    def examine = examine ? "--examine $examine" : ""
+    def extract = extract ? "--extract $extract" : ""
+    def skip = skip ? "--skip $skip" : ""
+    def type = csv ? "-c" : ""
+    def ambigous = examine ? "--ambigous $ambigous " : ""
+    def VERSION = '0.2.2.3' // WARN: Version information not provided by tool on CLI. Please update version string below when bumping container versions.
+    """
+    GMM-demux $type hto_files $hto_names $output_path $full_rep $n_cells $simplified_rep $examine $ambigous $extract $skip $args
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        GMM-Demux: $VERSION
+    END_VERSIONS
+    """
+
+    stub:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    def VERSION = '0.2.2.3'
+    """
+    mkdir test
+    touch test/barcodes.tsv.gz
+    touch test/features.tsv.gz
+    touch test/matrix.mtx.gz
+
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        GMM-Demux: $VERSION
+    END_VERSIONS
+    """
+}

modules/nf-core/gmmdemux/meta.yml

@@ -0,0 +1,110 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/meta-schema.json
+name: "gmmdemux"
+
+description: GMM-Demux is a Gaussian-Mixture-Model-based software for processing sample barcoding data (cell hashing and MULTI-seq).
+keywords:
+  - demultiplexing
+  - hashing-based deconvoltion
+  - single-cell
+tools:
+  - "gmmdemux":
+      description: "GMM-Demux is a Gaussian-Mixture-Model-based software for processing sample barcoding data (cell hashing and MULTI-seq)."
+      homepage: "https://pypi.org/project/GMM-Demux/"
+      documentation: "https://github.com/CHPGenetics/GMM-Demux"
+      tool_dev_url: "https://github.com/CHPGenetics/GMM-demux"
+      doi: "10.1186/s13059-020-02084-2"
+      licence: ["MIT"]
+
+input:
+  # Only when we have meta
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. `[ id:'sample1', single_end:false ]`
+  - cell_hashing_matrix:
+      type: file
+      description: path to file containing matrix from cell hashing data, the tool receives either CSV files or TSV, type must be specified using parameters
+  - cell_hashing_barcodes:
+      type: file
+      description: path to file containing barcodes from cell hashing data, the tool receives either CSV files or TSV, type must be specified using parameters
+  - cell_hashing_features:
+      type: file
+      description: path to file containing features from cell hashing data, the tool receives either CSV files or TSV, type must be specified using parameters
+  - csv:
+      type: string
+      description: |
+        Take input in csv format, instead of mmx format.
+  - hto_names:
+      type: string
+      description: |
+        Comma separated list of HTO names, without whitespace
+  - num_cells:
+      type: integer
+      description: |
+        Generate the statstic summary of the dataset. Requires an estimated total number of cells in the assay as input.
+  - output_dir:
+      type: string
+      description: |
+        The path for storing the Same-Sample-Droplets (SSDs).
+  - full_report:
+      type: string
+      description: |
+        Generate the full classification report. Require a path argument.
+  - simplified_report:
+      type: string
+      description: |
+        Generate the simplified classification report. Require a path argument.
+  - examine:
+      type: file
+      description: |
+        Provide the cell list. Requires a file argument. Only executes if num_cells is set
+  - skip:
+      type: string
+      description: |
+        Load a full classification report and skip the mtx folder as input. Require a path argument.
+  - extract:
+      type: string
+      description: |
+        Names of the sample barcoding tag(s) to extract, separated by ','. Joint tags are linked with '+'.
+  - ambigous:
+      type: integer
+      description: |
+        The estimated chance of having a phony GEM getting included in a pure type GEM cluster by the clustering algorithm.Only if exclude is executed
+output:
+  #Only when we have meta
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. `[ id:'sample1', single_end:false ]`
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - barcodes:
+      type: file
+      description: |
+        barcodes tsv file with removed cell-hashing-identifiable multiplets
+      pattern: "test/*.tsv.gz"
+  - matrix:
+      type: file
+      description: |
+        matrix mtx.tsv file with removed cell-hashing-identifiable multiplets
+      pattern: "test/*.tsv.gz"
+  - features:
+      type: file
+      description: |
+        features tsv file with removed cell-hashing-identifiable multiplets
+      pattern: "test/*.tsv.gz"
+authors:
+  - "@mari-ga"
+  - "@maxozo"
+  - "@wxicu"
+  - "@Zethson"
+maintainers:
+  - "@mari-ga"
+  - "@maxozo"
+  - "@wxicu"
+  - "@Zethson"

modules/nf-core/gmmdemux/nextflow.config

@@ -0,0 +1,10 @@
+params{
+    // Optional parameters for GMM-demux - Default values
+    threshold = 0.8
+}
+
+process {
+    withName: GMMDEMUX {
+        ext.args = " -t ${params.threshold} "
+    }
+}

modules/nf-core/gmmdemux/tests/main.nf.test

@@ -0,0 +1,86 @@
+// nf-core modules test gmmdemux
+nextflow_process {
+
+    name "Test Process GMMDEMUX"
+    script "../main.nf"
+    process "GMMDEMUX"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "gmmdemux"
+
+
+    test("Standard_Multiome - 10x mtx") {
+        when {
+            process {
+                """
+
+                input[0] = [
+                    [ id:'test'], 
+                    file(params.modules_testdata_base_path + "/genomics/homo_sapiens/10xgenomics/cellranger/hashing_demultiplexing/hto/barcodes.tsv.gz",checkIfExists: true),
+                    file(params.modules_testdata_base_path + "/genomics/homo_sapiens/10xgenomics/cellranger/hashing_demultiplexing/hto/matrix.mtx.gz",checkIfExists: true),
+                    file(params.modules_testdata_base_path + "/genomics/homo_sapiens/10xgenomics/cellranger/hashing_demultiplexing/hto/features.tsv.gz",checkIfExists: true),
+                    "MS-11,MS-12"
+                    ]
+                input[1] = ""
+                input[2] = ""
+                input[3] = "True"
+                input[4] = ""
+                input[5] = ""
+                input[6] = ""
+                input[7] = ""
+                input[8] = ""
+                input[9] = ""
+
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert path(process.out.barcodes.get(0).get(1)).exists() },
+
+            )
+        }
+
+    }
+
+    test("Standard_Multiome - 10x mtx - Stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+
+                input[0] = [
+                    [ id:'test'], 
+                    file(params.modules_testdata_base_path + "/genomics/homo_sapiens/10xgenomics/cellranger/hashing_demultiplexing/hto/barcodes.tsv.gz",checkIfExists: true),
+                    file(params.modules_testdata_base_path + "/genomics/homo_sapiens/10xgenomics/cellranger/hashing_demultiplexing/hto/matrix.mtx.gz",checkIfExists: true),
+                    file(params.modules_testdata_base_path + "/genomics/homo_sapiens/10xgenomics/cellranger/hashing_demultiplexing/hto/features.tsv.gz",checkIfExists: true),
+                    "MS-11,MS-12"
+                    ]
+                input[1] = ""
+                input[2] = ""
+                input[3] = ""
+                input[4] = ""
+                input[5] = ""
+                input[6] = ""
+                input[7] = ""
+                input[8] = ""
+                input[9] = ""
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert path(process.out.barcodes.get(0).get(1)).exists() },
+            )
+        }
+
+    }
+
+}

modules/nf-core/gmmdemux/tests/nextflow.config

@@ -0,0 +1,13 @@
+params{
+    // Optional parameters for GMM-demux - Default values
+    threshold = 0.8  
+
+}
+
+process {
+
+    withName: GMMDEMUX {
+        ext.args = " -t ${params.threshold} "
+    }
+
+}

modules/nf-core/gmmdemux/tests/tags.yml

@@ -0,0 +1,2 @@
+gmmdemux:
+  - "modules/nf-core/gmmdemux/**"